WARNINGPIT
A. Silovs couldn't make the stop
A. Silovs allowed 7 on 23 shots (0.696) — 4.61 goals worse than expected.
Reassess the crease rotation before the next start.
The verdict · stolen
The Red Wings stole one. They lost the expected-goals battle 2.39–3.60 and won anyway.
DET 7 · PIT 4
Preseason · exhibition
Split squads and prospects, not the club that will take the ice in October. The shots and grades below describe this game only: nothing here counts toward season records, form, rankings or player numbers.
◆ By the ModelSep 22, 9:41 PMxG: xgboost-0.758Edited for clarityDrafted from NHL play-by-play by Backcheck's model, worded for clarity by an AI editor that may not change a number. Nobody at Backcheck was at the rink. How this works
Pittsburgh outplayed Detroit by every expected goals measure — 60% of the workload in a ~2.70-goal league-average game — and lost 7-4. This is the defining split of the game: a Penguins team that built the better offense and tighter defense got run over by a Red Wings team that finished ruthlessly and received perfect relief.
Detroit's first goal, at 18:23 of the first, tied a shorthanded deficit. M. Brandsegg-Nygård's goal made it 1-1 after N. Robertson had given Pittsburgh an early advantage down a man. But Detroit took over in the second. M. Appleton scored at 04:26 to make it 2-1, then A. Sandin-Pellikka extended it to 3-1 thirty-one seconds later. R. Rakell cut it to 3-2 at 12:22, but the second period belonged to Detroit — they scored twice, Pittsburgh once, and by the start of the third the game was trending Detroit's way.
The third period was a coronation. M. Brandsegg-Nygård added a power-play goal at 05:37 to make it 4-2. K. Korczak pulled one back at 06:55 (4-3), but Detroit answered twice more — O. Becher at 09:24, W. Skoog at 12:39, J. Kiiskinen at 13:46. By the time R. McGroarty scored at 17:47 to make it 7-4, the Red Wings had already run away with it. A. Johansson's four assists in 22:27 traced the offense's heartbeat — he was on every critical goal, turning loose Red Wings into finished ones.
The expected goals told a very different story. Detroit generated 2.39 goals worth of chances; Pittsburgh generated 3.60. The Penguins controlled the higher-danger looks: 2–0 in scoring chances from the slot against Detroit's zero. On rush plays — where speed, not puck movement, creates opportunity — Pittsburgh's 32 chances to Detroit's 20 should have been decisive. Instead, Detroit converted 3 of those 20 into goals while Pittsburgh converted 4 of their 32. Finishing, not volume, was the difference.
Goaltending sealed Pittsburgh's defeat. A. Silovs faced 23 shots and allowed 7, posting a 0.696 save percentage that was 4.61 goals below expected against that workload. That is not a night where the team in front of you failed the math — that is a night where your goaltender handed the other team a three-goal head start on pure performance. Detroit's split wasn't clean. T. Augustine posted 6 saves on 8 shots (0.750 save percentage) running -1.10 versus expected, a weak night made invisible by the margin of victory. J. Gibson, in relief, saved 22 of 24 (0.917 save percentage) and ran +0.70 versus expected, which is workmanlike and all the buffer Detroit needed.
Pittsburgh built a night where they played the better game on paper, won the argument up and down the sheet, and still lost by three. That is the indignity of the scoreboard when the goaltender caves: the team in front was good, but it does not matter.